Table 1. Levels of evidence: classes of recommendation.
Definition | Phrasing | |||
---|---|---|---|---|
Level of evidence | ||||
A | Clear evidence for the recommendation | |||
Clearly proven through multicenter RCTs or meta-analysis with adequate content and power with high generalizability of findings | ||||
B | Reliable evidence for the recommendation | |||
Evidence found through well-performed cohort or patient-control group studies | ||||
C | Possible evidence for the recommendation | |||
Not reliable, but relevant evidence found through small RCTs, observational studies, or case series | ||||
E | Expert opinions | |||
No supporting evidence, but expert opinions based on clinical experience and expertise | ||||
Classes of recommendation | ||||
Class I | Clear evidence (A) and benefits, and high applicability in practice | Recommended | ||
Class IIa | Reliable evidence (B) and benefits, and high or moderate applicability in practice | Should be considered | ||
Class IIb | Unreliable evidence (C or D) and benefits, but high or moderate applicability in practice | May be considered | ||
Class III | Unreliable evidence (C or D), may cause harm, and low applicability in practice | Not recommended |
RCT, randomized controlled trial.